CX Methods
NPS misuse: when the score becomes a shield
An averaged NPS hides the journey stage driving customers out. Treat the score as a symptom to investigate, not a result to defend.
Net Promoter Score fails most companies not because the metric is broken but because of how it gets used. Scores get defended, rationalized, tied to bonuses, and averaged until the pain disappears from view. Each habit turns a useful signal into expensive noise.
NPS asks customers one question: how likely are you to recommend us? The score itself is harmless. The damage comes from what happens next. A low number gets explained away as a survey glitch or a rough quarter. A high number gets celebrated at the all-hands and shelved. Either way, nobody goes looking for root causes, and the score becomes a way to avoid fixing the experience rather than a reason to fix it. The most dangerous habit is averaging. Blend a brutal checkout experience with a pleasant onboarding and the company-wide score looks respectable, while the customers who hit the brutal stage are already gone.
Why it matters to the business
The score alone predicts less than most executives assume. A Journal of Marketing study by Keiningham and colleagues, covering 21 firms and more than 15,500 customer interviews, found NPS performed no better than ordinary satisfaction indices at predicting revenue growth. Even the metric's creator agrees it gets abused. Fred Reichheld wrote in Harvard Business Review that linking scores to frontline bonuses breeds pleading, score-chasing, and gaming, and proposed an audited, accounting-based Earned Growth Rate as the hard complement.
Misuse also blinds leaders to churn. Bain research found 60 to 80 percent of lost customers described themselves as satisfied or very satisfied on surveys shortly before defecting. A defended score and a churning customer base can coexist comfortably for years.
How to use it
- Report NPS by journey stage, never as one blended company average.
- Decouple the score from individual bonuses so nobody is paid to plead for tens.
- Require a root-cause investigation for every material score movement, up or down.
- Validate the score against repeat revenue and net revenue retention each quarter.
- Audit sampling and response bias annually; a gamed sample is worse than no score.
Where teams get it wrong
The classic failure is celebration without curiosity. A strong score gets treated as proof the experience is fine, so investment shifts elsewhere, and the stage-level pain the average concealed keeps quietly feeding churn. By the time the blended number finally moves, the damage is quarters old and the customers are unrecoverable.
Ask your team
- Can we see NPS broken out by journey stage, or only one blended number?
- Is any bonus in this company tied to a survey score, and how would we detect gaming?
- When the score last moved, what root cause did we find and what did we ship?
A score you defend is a score you have stopped learning from.
Apply this
Reading about nps misuse: when the score becomes a shield is one thing. Seeing where it applies in your journey is the useful part.